Erythropolesis, the production of red blood cells (RBCs), is regulated by an intricate network of signals involving the hormone erythropoietin (Epo), and sensing of hypoxia through inhibition of both HIF-1alpha proline hydroxylation and its subsequent interaction with the Von Hippel-Lindau (VHL) protein. Either insufficient or excess erythropoiesis produces disease in humans, with too few RBCs resulting in anemia, and too many RBCs resulting in polycythemia vera. The treatment of anemia alone consumes significant economic resources, with the necessity for either blood transfusions or Epo treatment. We have developed adenoviruses which express soluble ectodomains of the Flkl and Fltl receptors for Vascular Endothelial Growth Factor (VEGF). Single injections in mice of these adenoviruses produce high levels of circulating soluble Flk1 and Flt1 ectodomains which persist for >3-4 weeks, potently suppressing tumor growth and angiogenesis, and producing stringent conditional inactivation of-VEGF in adult animals. Surprisingly, adenoviral expression of Flk1 or Flt1 ectodomains (soluble VEGF receptors, or sVEGFRs) in normal mice strongly stimulates erythropoiesis, with increase in hematocrit increasing from baseline 45% to new levels of 55-75 %. Circulating erythropoietin levels are elevated in parallel; however, the Epo originates from the liver as in embryonic development, not the usual site of synthesis in the kidney. This data implicates VEGF as an unsuspected repressor of hepatic Epo synthesis, and this grant application is focused on determining the underlying moIecular mechanisms. In Aim 1, experiments will expand upon preliminary data implicating perturbation of hepatocyte--endothelial cross-talk in the sVEGFR induction of hepatic Epo. Adenoviral and transgenic expression of Cre recombinase will be used to test the effects of specific inhibition of hepatocyte-produced VEGF on hepatocyte Epo expression. Conversely, co-culture experiments manipulating VEGF will be used to test the effects of endothelial cells on hepatocyte Epo production. Aim 2 will define the specific factors mediating the transcriptional regulation of Epo by VEGF and VEGF blockade. Candidate factors (HIF proteins, HNF4, RXRalpha, GATA factors) will be directly tested by chromatin immunoprecipitation assays and functionally by adenoviral and transgenic expression of Cre to delete floxed alleles of relevant candidates. Additionally, novel factors will be sought by unbiased approaches of DNAse1 footprinting and hypersensitivity assays. The therapeutic implications of the ability to reactivate hepatic Epo synthesis are substantial. These studies should also yield significant insight into the action of VEGF as an unsuspected upstream regulator of RBC homeostasis, into the regulation of hepatic Epo production which is strongly repressed post-natally, and indicate the potential utility of alterations in erythropoiesis as surrogate markers for or a desirable consequence of stringent VEGF blockade.